Lund University Global Drought Dataset (LUGD), 1981–2025: Weekly 0.1° saSPEI with Accumulation Timescales from 1 Week to 48 Months – 90N-45N

Dataset description Note the dataset is hosted at: https://zenodo.org/records/21810474 under Global drought studies community This dataset provides a snow-adjusted Standardized Precipitation Evapotranspiration Index (saSPEI) for northern terrestrial areas between 45° N and 90° N. The dataset is provided on a regular 0.1° latitude–longitude grid, at nominal weekly step (four nominal time steps per month), for the period 1981–2025. Twelve accumulation timescales are included: 1 and 2 weeks, and 1, 2, 3, 6, 9, 12, 18, 24, 36, and 48 months. A corresponding conventional SPEI dataset, generated using the same meteorological forcing, spatial grid, temporal calendar, calibration period, accumulation procedure, probability distribution, and temporal alignment, is available separately at https://zenodo.org/records/18985819. The two products can therefore be compared directly to quantify the effects of explicitly accounting for seasonal snow storage and release. Purpose and conceptual framework Conventional SPEI is calculated from the climatic water balance between precipitation and potential evapotranspiration. In cold regions, however, a substantial fraction of precipitation may fall as snow and remain stored in the snowpack for days to months before becoming available at the land surface. Treating snowfall as immediately available water can therefore misrepresent the timing and magnitude of climatic water availability. The saSPEI addresses this limitation by replacing total precipitation with the effective liquid-water flux released from the snow–rain system. The snow-adjusted climatic water balance is defined as: D = Q − PET where Q is the effective precipitation reaching the land surface and PET is potential evapotranspiration. For the conventional SPEI product, the corresponding balance is: D = P − PET where P is total precipitation. Meteorological forcing The dataset was generated using the following meteorological inputs: precipitation from the Multi-Source Weighted-Ensemble Precipitation dataset, MSWEP; potential evapotranspiration calculated using the Bristol potential evapotranspiration formulation; 2-m air temperature from ERA5-Land; and 10-m wind speed derived from ERA5-Land wind components. All forcing data were harmonized to the common 0.1° spatial grid and daily temporal resolution before calculation of the snow-adjusted water balance. Rain–snow partitioning Daily precipitation is partitioned into rainfall and snowfall using a temperature-dependent seasonal rain–snow partitioning scheme based on the snowfall-probability formulation of Dai (2008). The method represents the gradual transition between rain and snow rather than applying a single fixed temperature threshold. Rainfall is made directly available to the land-surface water balance, subject to the treatment of temporary liquid-water storage within the snowpack. Snowfall is added to the solid snow reservoir and becomes available only after melt and subsequent release from the snowpack. Snow model Snow accumulation, internal storage, and liquid-water release are simulated using a one-layer bulk thermodynamic snow model adapted from the snow-process representation in LPJ-GUESS 4.2 (Pongracz, A., et al. 2021). The model represents: solid snow-water storage; liquid water retained within the snowpack; fresh-snow density; snow densification and compaction; bulk snow temperature; snow cold content; temperature- and precipitation-dependent snowmelt; refreezing of liquid water; rain-on-snow inputs; and liquid-water overflow from the snowpack. The effective precipitation flux, Q, consists of liquid water released to the land surface through rainfall, meltwater, rain-on-snow bypass, and overflow from snowpack liquid storage, after accounting for refreezing and temporary retention within the snowpack. This formulation shifts the contribution of snowfall from the date of precipitation to the date on which the water is physically released from the snowpack. Temporal aggregation and accumulation Daily climatic water-balance values are aggregated into four nominal weekly step per month: day 1; day 9; day 16; and day 23. Statistical standardization For each accumulation timescale, accumulated climatic water-balance values are standardized separately for each of the 48 nominal calendar groups. A three-parameter log-logistic probability distribution is fitted to the accumulated water-balance series for each grid cell and calendar group over the 1981–2025 calibration period. Distribution parameters are estimated using probability-weighted moments following the SPEI methodology of Vicente-Serrano et al. (2010). The fitted cumulative probabilities are transformed to standard normal variates. The resulting saSPEI values are dimensionless and approximately normally distributed within each calendar group. Negative values indicate drier-than-normal conditions, whereas positive values indicate wetter-than-normal conditions for the corresponding location, calendar period, and accumulation timescale. Values close to zero indicate conditions near the 1981–2025 climatological median. File structure The dataset is distributed as NetCDF files containing CF-style time, latitude, and longitude coordinates. The principal data variable is: saSPEI: snow-adjusted Standardized Precipitation Evapotranspiration Index. The separately archived conventional product contains: SPEI: conventional Standardized Precipitation Evapotranspiration Index, separately at https://zenodo.org/records/18985819 due to limit of Zenodo. Both variables are dimensionless. Individual NetCDF files correspond to specific accumulation timescales. Recommended applications The dataset is intended for: drought monitoring and climatological drought assessment at northern mid- and high latitudes; analysis of drought impacts on vegetation, ecosystems, carbon uptake, and soil-water availability; investigation of drought timing in seasonally snow-covered regions; comparison between conventional and snow-adjusted drought characterization; analysis of long-term changes in drought frequency, duration, severity, and spatial extent; and evaluation of climate-model or reanalysis-based drought indicators. Scope and limitations The saSPEI represents meteorological water availability after accounting for the temporary storage and delayed release of precipitation by a seasonal snowpack. It is not a complete hydrological or land-surface model. The dataset does not explicitly represent: frozen-soil processes or infiltration limitations; soil-moisture storage and vertical soil-water transport; groundwater storage; glacier or ice-sheet mass balance; wind-driven snow redistribution; sublimation from blowing snow; sub-grid topographic variation; lateral surface or subsurface flow; river and catchment routing; or water-management effects. The index should therefore be interpreted as a standardized climatic drought indicator rather than a direct estimate of soil moisture, runoff, streamflow, groundwater, or ecosystem stress. Its reliability may also be affected by uncertainties in the meteorological forcing, potential evapotranspiration estimates, rain–snow partitioning, snow-model parameters, and probability-distribution fitting. Files